Impact of Representations in Assessments on Student Performance and Equity

Calls for assessments incorporating representations beyond the symbolic level (e.g., chemical reactions and formulas) have encouraged assessment designers to choose from a variety of representations in the design of chemistry assessments. This work expands on prior work in considering how representations are incorporated within assessments. First, students were given assessment items similar in task with common rationales for the design of answer choices with different representations substituted across the items. Second, students were randomly assigned to items identical in prompt, answer choices, and symbolic representation with experimental items supplementing symbolic representation with particulate or mathematical representation. The impact was operationalized as student performance, the performance of at-risk chemistry students scoring in the bottom quartile of the cohort's standardized math test scores, and on the answer choices to which students attend. Substituting symbolic for particulate representation had a significant, negative impact on overall and at-risk student performance and a considerable impact on students' selection of distractors. However, supplementing symbolic with a particulate representation had no discernible impact. Supplementing symbolic representations with tables including a column for "moles" had a significant, positive impact on overall student performance with considerable improvement observed among at-risk students. Students' selection of a distractor representative of a failure to apply a mole-to-mole ratio was reduced when symbolic representations were supplemented with tables. Practical implications for the use of representations in a more equitable assessment design, and as a learning experience for students overall, are described.